PilotScope is a middleware to bridge the gaps of deploying AI4DB (Artificial Intelligence for Databases) algorithms into actual database systems.

Dynatrace hash library for Java

Fast HyperLogLog for Python.

Paper related to AI4DB techniques

Implementation of DeepDB: Learn from Data, not from Queries!

Neural Relation Understanding: neural cardinality estimators for tabular data

State-of-the-art neural cardinality estimators for join queries

Estimating k-mer coverage histogram of genomics data

Rust's fastest and most accurate cardinality estimators.

Paper about the estimation of cardinalities from HyperLogLog sketches

SetSketch: Filling the Gap between MinHash and HyperLogLog

A crate for estimating the cardinality of distinct elements in a stream or dataset.

A Unified Deep Model of Learning from both Data and Queries for Cardinality Estimation

No description provided.

A pytorch implementation for FACE: A Normalizing Flow based Cardinality Estimator

A Python library for efficient feature ranking and selection on sparse data sets.

Fast and Memory Efficient Genome Sketching via HyperLogLog, HyperMinHash and UltraLogLog

Fast Cardinality Estimation of Multi-Join Queries Using Sketches

ExaLogLog: Space-Efficient and Practical Approximate Distinct Counting up to the Exa-Scale

Code for Local Deep Learning Models for Cardinality Estimation

UltraLogLog: A Practical and More Space-Efficient Alternative to HyperLogLog for Approximate Distinct Counting

[VLDB'22] Cardinality Estimation of Approximate Substring Queries using Deep Learning.

An implementation of the algorithms presented in the paper "Cardinality Estimation Done Right: Index-Based Join Sampling"

Probabilistic data structures (bloom filter / counting bloom filter / linear counter)

Learned model to estimate number of distinct values (NDV) of a population using a small sample.

Datasets Used in AI4DB Research Work

Real-time cardinality estimator built for the Future of Database Programming Contest. Implements histogram-based methods, HyperLogLog, MCV tracking, and optimized data structures for high-performance. Achieved first place in the competition

Code for variable skipping ICML 2020 paper

Scalable Join Cardinality Estimaitor

Simpli-Squared is a statistics-free join ordering algorithm Without Cardinality Estimates.